In India, where despite the growing population of tigers, poaching is proliferating, a combination of data analytics and AI tools that collect and process wildlife crime data and predict tiger poaching, will change the way tiger reserves and conservation efforts work. 

How AI can help protect India’s tigers

For an animal that requires no human intervention to thrive, yet needs human protection, technology may be the only path to conservation

Souma Das, Managing Director, Teradata India

Tigers are one of the most majestic and widely recognised big cats across the world. With a dwindling population of only 3,8901 across the globe, these apex predators stick to the tropical rainforests, evergreen forests, mangrove swamps, grasslands and savannahs of Asia. Though they are the national animal for India, Bangladesh, Malaysia, South Korea, Myanmar and Vietnam; tigers have lost 93 per cent2 of their historical range due to the destruction of their habitat, human-wildlife conflict, climate change and poaching. Tiger population has plummeted 95 per cent3 in the last century, putting the animal on the global endangered list, with some sub-species extinct in the wild.

In 2006, the tiger population had declined to 1,411, making national and international agencies launch tiger conservation, anti-poaching and sensitisation campaigns. Being home to 60 per cent4 of the world tiger population, India launched several campaigns such as Save the Tiger by NDTV-Aircel, Save our Tigers by Wildlife Conservation Trust, and they boosted efforts of the 1973 Project Tiger conservation program by the government. However, since 2012, 40 per cent5 of tiger deaths in India have been attributed to poaching, raising questions as to how best to protect the country’s national animal?

The answer: Technology

Demarcating tiger reserves, ramping up forest security and sensitisation campaigns for the population are not the only methods to save tigers. For an animal that requires no human intervention to thrive, yet needs human protection, technology may be the only path to conservation. For example, in India, the National Tiger Conservation Authority brought technology into the mix by launching projects such as ‘M-Stripes’ and ‘E-Eye’6.

The M-Stripes system integrates ecological insights from tiger states with GIS tools to assess the intensity and spatial coverage of patrols by producing reports and maps that synthesize information on illegal activities, wildlife crime, protection efforts, ecological status. The ‘E-Eye’ on the other hand is a pilot being executed in Corbett Tiger Reserve that consists of inter-connected high-mounted short-range infra-red cameras that detect movement of anything above 20 Kilograms in a 300-square-mile radius. On detecting anomalies, it alerts the field stations for appropriate action and has become a deterrent to poaching.

Having said that, poachers have also evolved. They have mapped patrolling routes, avoid regular trails, they know where the cameras are, they can anticipate animal movement and have resources to track their prey. They have advanced weapons as well as know how to circumvent laws to stay protected. What else can be done to counter poaching?

AI the saviour

Aerial-drones, infra-red cameras, real-time monitoring devices, RFID tags and GPS geo-location for surveillance are being used for wildlife conservation across the world. However, these technologies work more on the lines of collating huge volumes of data just for visualisation. Signs of poaching, wildlife observations, wildlife criminal arrests, targeted animal movements and other forest patrol results can be fed into big data analytics platforms to be processed into charts, maps and reports that can be easily analysed. If AI, image processing and predictive analysis are brought into the picture, it will help standardise monitoring, enforcement and increase efficiency.

For example, researchers from the University of Southern California Centre for Artificial Intelligence in Society7 have developed a tool to help catch poachers in near real-time called SPOT or Systematic Poacher Detector. Since infrared cameras are sometimes unable to distinguish between the heat of humans and animals, the researchers fed 180,000 human and animal heatmaps captured by the cameras to a deep learning algorithm to identify humans. The tech was set up at the national parks in Zimbabwe and Malawi, with AI processing live feeds from infrared drone cameras, that was able to spot poachers and animals in three-tenths of a second.

Another AI innovation that could help India in its tiger conservation and the anti-tiger-poaching campaign is PAWS or Protection Assistant for Wildlife Security. Developed five years back and tested in Malaysia and Uganda, the AI application incorporates game theory model for wildlife protection. It analyses terrain and topography, incorporates paths of the animals and patrol routes that poachers avoid and applies it to a machine learning algorithm that prescribes randomised patrol routes to predict routes poachers might take.

Last year, wildlife conservationists and forest patrol combined PAWS with another algorithm called CAPTURE8 or Comprehensive Anti-Poaching Tool with Temporal and Observation Uncertainty Reasoning. CAPTURE uses similar data and tools like PAWS and SPOT to predict the area and likeliness of an attack rather than predictive route mapping.

In India, where despite the growing population of tigers, poaching is proliferating, a combination of data analytics and AI tools that collect and process wildlife crime data and predict tiger poaching, will change the way tiger reserves and conservation efforts work. Those reserves that have had experience with infrared cameras and drones, as well as those which are understaffed, underequipped and poorly armed can use AI and analytics to make anti-poaching futureproof.